‘The Bell Curve: Intelligence and Class Structure in American Life’ by Richard Herrnstein and Charles Murray (New York: Free Press, 1994).
There’s no such thing as bad publicity’ – or so contends a famous adage of the marketing industry.
‘The Bell Curve: Intelligence and Class Structure in America’ by Richard Herrnstein and Charles Murray is perhaps a case in point.
This dry, technical, academic social science treatise, full of statistical analyses, graphs, tables, endnotes and appendices, and totalling almost 900 pages, became an unlikely nonfiction bestseller in the mid-1990s on a wave of almost universally bad publicity in which the work was variously denounced as racist, pseudoscientific, fascist, social Darwinist, eugenicist and sometimes even just plain wrong.
Readers who hurried to the local bookstore eagerly anticipating an incendiary racialist polemic were, however, in for a disappointment.
As a result, as with that other book, many myths have emerged regarded the content of ‘The Bell Curve’ that are quite contradicted when one actually takes the time and trouble to read it for oneself.
Thus, ‘The Bell Curve’ is a book so controversial that the controversy begins with the very identification of its subject-matter.
On the one hand, the book’s critics focused almost exclusively on subject of race. This led to the common perception that ‘The Bell Curve’ was a book about race and race differences in intelligence.
Ironically, many racialists seem to have taken these leftist critics at their word, enthusiastically citing the work as support for their own views regarding race differences in intelligence.
On the other hand, however, surviving co-author Charles Murray insisted from the outset that the issue of race, and race differences in intelligence, was always peripheral to he and co-author Richard Herrnstein’s primary interest and focus, which was, he claimed, on the supposed emergence of a ‘Cognitive Elite’ in modern America.
Actually, however, both these views seem to be incorrect. While the first section of the book does indeed focus on the supposed emergence of a ‘Cognitive Elite’ in modern America, the overall theme of the book seems to be rather broader.
Thus, the second section of the book focuses on the association between intelligence and various perceived social pathologies, such as unemployment, welfare dependency, illegitimacy, crime and single-parenthood.
To the extent the book has a single overarching theme, one might say that it is a book about the social and economic correlates of intelligence, as measured by IQ tests, in modern America.
Its overall conclusion is that intelligence is indeed a strong predictor of social and economic outcomes for modern Americans – high intelligence with socially desirable outcomes and low intelligence with socially undesirable ones.
On the other hand, however, the topic of race is not quite as peripheral to the book’s themes as sometimes implied by Murray and others.
Thus, it is sometimes claimed only a single chapter dealt with race. Actually, however, two chapters focus on race differences, namely chapters 13 and 14, respectively titled ‘Ethnic Differences in Cognitive Ability’ and ‘Ethnic Inequalities in Relation to IQ’.
In addition, a further two chapters, namely chapters 19 and 20, entitled respectively ‘Affirmative Action in Higher Education’ and ‘Affirmative Action in the Workplace’, deal with the topic of affirmative action, as does the final appendix, entitled ‘The Evolution of Affirmative Action in the Workplace’ – and, although affirmative action has been employed to favour women as well as racial minorities, it is with racial preferences that Herrnstein and Murray are primarily concerned.
However, these chapters represent only 142 of the book’s nearly 900 pages.
Moreover, in much of the remainder of the book, the authors actually explicitly restrict their analysis to white Americans exclusively. They do so precisely because the well documented differences between the races in IQ as well as in many of the social outcomes whose correlation with IQ the book discusses would mean that race would have represented a potential confounding factor that they would otherwise have to take steps to control for.
Herrnstein and Murray therefore took to decision to extend their analysis to race differences near the end of their book, in order to address the question of the extent to which differences in intelligence, which they have already demonstrated to be an important correlate of social and economic outcomes among whites, are also capable of explaining differences in achievement as between races.
Race and Intelligence
If the first controversy of ‘The Bell Curve’ concerns whether it is a book primarily about race and race differences in intelligence, the second controversy is over what exactly the authors concluded with respect to this vexed and contentious issue.
Thus, the same leftist critics who claimed that ‘The Bell Curve’ was primarily a book about race and race differences in intelligence, also accused the authors of concluding that black people are innately less intelligent than whites.
Some racists, as I have already noted, evidently took the leftists at their word, and enthusiastically cite the book as support and authority for this view.
However, in subsequent interviews, Murray always insisted he and Herrnstein had actually remained “resolutely agnostic” on the extent to which genetic factors underlay the IQ gap.
In the text itself, Herrnstein and Murray do indeed declare themselves “resolutely agnostic” with regard to the extent of the genetic contribution to the test score gap (p311).
However, just couple of sentences before they use this very phrase, they also appear to conclude that genes are indeed at least part of the explanation, writing:
“It seems highly likely to us that both genes and the environment have something to do with racial differences [in IQ]” (p311).
This paragraph, buried near the end of chapter 13, during an extended discussion of evidence relating to the causes of race differences in intelligence, is the closest the authors come to actually declaring any definitive conclusion regarding the causes of the black-white test score gap.
This conclusion, though phrased in sober and restrained terms, is, of course, itself sufficient to place its authors outside the bounds of acceptable opinion in the early-twenty-first century, or indeed in the late-twentieth century when the book was first published, and is sufficient to explain, and, for some, justify, the opprobrium heaped upon the book’s surviving co-author from that day forth.
Intelligence and Social Class
It seems likely that races which evolved on separate continents, in sufficient reproductive isolation from one another to have evolved the obvious (and not so obvious) physiological differences between races that we all observe when we look at the faces, or bodily statures, of people of different races (and that we indirectly observe when we look at the results of different athletic events at the Olympic Games), would also have evolved to differ in psychological traits, including intelligence.
Indeed, it is surely unlikely, on a priori grounds alone, that all different human races have evolved, purely by chance, the exact same level of intelligence.
However, if races differ in intelligence are therefore probable, the case for differences in intelligence as between social classes is positively compelling.
Indeed, on a priori grounds alone, it is inevitable that social classes will come to differ in IQ, if one accepts two premises, namely:
1) Increased intelligence is associated with upward social mobility; and
2) Intelligence is passed down in families.
In other words, if more intelligent people tend, on average, to get higher-paying jobs than those of lower intelligence, and the intelligence of parents is passed on to their offspring, then it is inevitable that the offspring of people with higher-paying jobs will, on average, themselves be of higher intelligence than are the offspring of people with lower paying jobs.
Incidentally, this second premise, namely that intelligence is passed down in families, does not depend on the heritability of IQ in the strict biological sense. After all, even if heritability of intelligence were zero, intelligence could still be passed down in families by environmental factors (e.g. the ‘better’ parenting techniques of high IQ parents, or the superior material conditions in wealthy homes).
The existence of an association between social class and IQ ought, then, to be entirely uncontroversial to anyone who takes any time whatsoever to think about the issue.
If there remains any room for reasoned disagreement, it is only over the direction of causation – namely the question of whether:
1) High intelligence causes upward social mobility; or
2) A privileged upbringing causes higher intelligence.
These two processes are, of course, not mutually exclusive. Indeed, it would seem intuitively probable that both factors would be at work.
Interestingly, however, evidence demonstrates the occurrence only of the former.
Thus, even among siblings from the same family, the sibling with the higher childhood IQ will, on average, achieve higher socioeconomic status as an adult. Likewise, the socioeconomic status a person achieves as an adult correlates more strongly with their own IQ score than it does with the socioeconomic status of their parents or of the household they grew up in (see Straight Talk About Mental Tests: p195).
In contrast, family, twin and adoption studies and of the sort conducted by behavioural geneticists have concurred in suggesting that the so-called shared family environment (i.e. those aspects of the family environment shared by siblings from the same household, including social class) has but little effect on adult IQ.
In other words, children raised in the same home, whether full- or half-siblings or adoptees, are, by the time they reach adulthood, no more similar to one another in IQ than are children of the same degree of biological relatedness brought up in entirely different family homes (see The Nurture Assumption: reviewed here).
However, while the direction of causation may still be disputed by intelligent (if uninformed) laypeople, the existence of an association between intelligence and social class ought not, one might think, be in dispute.
However, in Britain today, in discussions of social mobility, if children from deprived backgrounds are underrepresented, say, at elite universities, then this is almost invariably taken as incontrovertible proof that the system is rigged against them. The fact that children from different socio-economic backgrounds differ in intelligence is almost invariably ignored.
For his trouble, Charlton found himself condemned by the National Union of Students and assorted rent-a-quote academics and professional damned fools, while even the ostensibly ‘right-wing’ Daily Mail newspaper saw fit to publish a headline “Higher social classes have significantly HIGHER IQs than working class, claims academic”, as if this were in some way a controversial or contentious claim (Clark 2008).
Meanwhile, when, in the same year, a professor at University College a similar point with regard the admission of working-class students to medical schools, even the then government Health Minister, Ben Bradshaw, saw fit to offer his two cents worth (which were not worth even that), declaring:
“It is extraordinary to equate intellectual ability with social class” (Beckford 2008).
Actually, however, what is truly extraordinary is that any intelligent person, least of all a government minister, would dispute the existence of such a link.
Herrnstein’s syllogism leads to a related paradox – namely that, as environmental conditions are equalized, heritability increases.
Thus, as large differences in the sorts of environmental factors known to affect IQ (e.g. malnutrition) are eliminated, so differences in income have come to increasingly reflect differences in innate ability.
Moreover, the more gifted children from deprived backgrounds who escape their humble origins, then, given the substantial heritability of IQ, the fewer such children will remain among the working-class in subsequent generations.
The result is what Herrnstein and Murray call the ‘Cognitive Stratification’ of society and the emergence of what they call a ‘Cognitive Elite’.
Thus, in feudal society, a man’s social status was determined largely by ‘accident of birth’ (i.e. he inherited the social station of his father).
However, today, a person’s social status, at least according to Herrnstein and Murray, is determined primarily, and increasingly, by their level of intelligence.
Of course, people are not allocated to a particular social class by IQ testing itself. Indeed, the use of IQ tests by employers and educators has been largely outlawed on account of its ‘disparate impact‘ (or ‘indirect discrimination’, to use the equivalent British phrase) with regard to race (see below).
However, the skills and abilities increasingly valued at a premium in western society (and, increasingly, many non-western societies as well), mean that, through the operation of the education system and labour market, individuals are effectively sorted by IQ, even without anyone ever actually sitting an IQ test.
In other words, society is becoming increasingly meritocratic – and the form of ostensible ‘merit’ upon which attainment is based is intelligence.
For Herrnstein and Murray, this is a mixed blessing:
“That the brightest are identified has its benefits. That they become so isolated and inbred has its costs” (p25).
However, the correlation between socioeconomic status and intelligence remains imperfect.
For one thing, there are still a few highly remunerated, and very high-status, occupations that rely on skills that are not especially, if at all, related to intelligence. I think here, in particular, of professional sports and the entertainment industry. Thus, leadings actors, pop stars and sports stars are sometimes extremely well-remunerated, and very high-status, but may not be especially intelligent.
More importantly, while highly intelligent people might be, by very definition, the only ones capable of performing cognitively-demanding, and hence highly remunerated, occupations, this is not to say all highly intelligent people are necessarily employed in such occupations.
Thus, whereas all people employed in cognitively-demanding occupations are, almost by definition, of high intelligence, people of all intelligence levels are capable of doing cognitively-undemanding jobs.
Thus, a few people of high intellectual ability remain in low-paid work, whether on account of personality factors (e.g. laziness), mental illness, lack of opportunity or sometimes even by choice (which choice is, of course, itself a reflection of personality factors).
Therefore, the correlation between IQ and occupation is far from perfect.
The sorting of people with respect to their intelligence begins in the education system. However, it continues in the workplace.
Thus, general intelligence, as measured by IQ testing, is, the authors claim, the strongest predictor of occupational performance in virtually every occupation. Moreover, in general, the higher paid and higher status the occupation in question, the stronger the correlation between performance and IQ.
However, Herrnstein and Murray are at pains to emphasize, intelligence is a strong predictor of occupational performance even in apparently cognitively undemanding occupations, and indeed almost always a better predictor of performance than tests of the specific abilities the job involves on a daily basis.
However, in the USA, employers are barred from using testing to select among candidates for a job or for promotion unless they can show the test has ‘manifest relationship’ to the work, and the burden of proof is on the employer to show such a relationship. Otherwise, given their ‘disparate impact’ with regard to race (i.e. the fact that some groups perform worse), the tests in question are deemed indirectly discriminatory and hence unlawful.
Therefore, employers are compelled to test, not general ability, but rather the specific skills required in the job in question, where a ‘manifest relationship’ is easier to demonstrate in court.
However, since even tests of specific abilities almost invariably still tap into the general factor of intelligence, races inevitably score differently even on these tests.
Indeed, because of the ubiquity and predictive power of the g factor, it is almost impossible to design any type of standardized test, whether of specific or general ability or knowledge, in which different racial groups do not perform differently.
However, if some groups outperform others, the American legal system presumes a priori that this reflects test bias rather than differences in ability.
Therefore, although the words “all men are created equal” are not, contrary to popular opinion, part of the US constitution, the Supreme Court has effectively decided, by legal fiat, to decide cases as if they were.
However, just as a law passed by Congress cannot repeal the law of gravity, so a legal presumption that groups are equal in ability cannot make it so.
Thus, the bar on the use of IQ testing by employers has not prevented society in general from being increasingly stratified by intelligence, the precise thing measured by the outlawed tests.
Nevertheless, Herrnstein and Murray estimate that the effective bar on the use of IQ testing makes this process less efficient, and cost the economy somewhere between 80 billion to 13 billion dollars in 1980 alone (p85).
Conscientiousness and Career Success
I am skeptical of Herrnstein and Murray’s conclusion that IQ is the best predictor of academic and career success. I suspect hard work, not to mention a willingness to toady, toe the line, and obey orders, is at least as important in even the most cognitively-demanding careers, as well as in schoolwork and academic advancement.
Perhaps the reason these factors have not (yet) been found to be as highly correlated with earnings as is IQ is that we have not yet developed a way of measuring these aspects of personality as accurately as we can measure a person’s intelligence through an IQ test.
For example, the closest psychometricians have come to measuring capacity for hard work is the personality factor known as ‘conscientiousness’, one of the Big Five factors of personality revealed by psychometric testing.
Conscientiousness does indeed correlate with success in education and work (e.g. Barrick & Mount 1991). However, the correlation is weaker than that between IQ and success in education and at work.
However, this may be because personality is less easily measured by current psychometric methods than is intelligence – not least because personality tests generally rely on self-report, rather than measuring actual behaviour.
Thus, to assess conscientiousness, questionnaires ask respondents whether they ‘see themselves as organized’, ‘as able to follow an objective through to completion’, ‘as a reliable worker’, etc.
This would be the equivalent of an IQ test that, instead of directly testing a person’s ability to recognize patterns or manipulate shapes by having them do just this, simply asked respondents how good they perceived themselves as being at recognizing patterns, or manipulating shapes.
Obviously, this would be a less accurate measure of intelligence than a normal IQ test. After all, some people lie, some are falsely modest and some are genuinely deluded.
Indeed, according to the Dunning Kruger effect, it is those most lacking in ability who most overestimate their abilities – precisely because they lack the ability to accurately assess their ability (Kruger & Dunning 1999).
In an IQ test, on the other hand, one can sometimes pretend to be dumber than one is, by deliberately getting questions wrong that one knows the answer to.
However, it is not usually possible to pretend to be smarter than one is by getting more questions right simply because one would not know what are the right answers.
‘Affirmative Action’ and Test Bias
In chapters nineteen and twenty, respectively entitled ‘Affirmative Action in Higher Education’ and ‘Affirmative Action in the Workplace’, the authors discuss so-called ‘affirmative action’, an American euphemism for systematic and overt discrimination against white males.
It is well-documented that, in the United States, blacks, on average, earn less than white Americans. On the other hand, it is less well-documented that whites, on average, earn less than people of Indian, Chinese and Jewish ancestry.
With the possible exception of Indian-Americans, these differences, of course, broadly mirror those in average IQ scores.
Indeed, according to Herrnstein and Murray, the difference in earnings between whites and blacks, not only disappears after controlling for differences in IQ, but is actually partially reversed. Thus, blacks are actually somewhat overrepresented in professional and white-collar occupations as compared to whites of equivalent IQ.
This remarkable finding Herrnstein and Murray attribute to the effects of affirmative action programmes, as black Americans are appointed and promoted beyond what their ability merits because through discrimination.
There, they concluded that testing was not biased against African-Americans, because, among other reasons, IQ tests were equally predictive of real-world outcomes (e.g. in education and employment) for both blacks and whites, and blacks do not perform any better in the workplace or in education than their IQ scores predict.
This is, one might argue, not wholly convincing evidence that IQ tests are not biased against blacks. It might simply suggest that society at large, including the education system and the workplace, is just as biased against blacks as are the hated IQ tests. This is, of course, precisely what we are often told by the television, media and political commentators who insist that America is a racist society, in which such mysterious forces as ‘systemic racism’ and ‘white privilege’ are pervasive.
In fact, the authors acknowledge this objection, conceding:
“The tests may be biased against disadvantaged groups, but the traces of bias are invisible because the bias permeates all areas of the group’s performance. Accordingly, it would be as useless to look for evidence of test bias as it would be for Einstein’s imaginary person traveling near the speed of light to try to determine whether time has slowed. Einstein’s traveler has no clock that exists independent of his space-time context. In assessing test bias, we would have no test or criterion measure that exists independent of this culture and its history. This form of bias would pervade everything” (p285).
Herrnstein and Murray ultimately reject this conclusion on the grounds that it is simply implausible to assume that:
“[So] many of the performance yardsticks in the society at large are not only biased, they are all so similar in the degree to which they distort the truth-in every occupation, every type of educational institution, every achievement measure, every performance measure-that no differential distortion is picked up by the data” (p285).
In fact, however, Nicholas Mackintosh identifies one area where IQ tests do indeed under-predict black performance, namely with regard to so-called adaptive behaviours – i.e. the ability to cope with day-to-day life (e.g. feed, dress, clean, interact with others in a ‘normal’ manner).
Yet Herrnstein and Murray seem to have inadvertently, and evidently without realizing it, identified yet another sphere where standardized testing does indeed under-predict real-world outcomes for blacks.
Thus, if indeed, as Herrnstein and Murray claim, blacks are somewhat overrepresented among professional and white-collar occupations relative to their IQs, this suggests that blacks do indeed do better in real-world outcomes than their test results would predict and, while Herrnstein and Murray attribute this to the effect of discrimination against whites, it could instead surely be interpreted as evidence that the tests are biased against blacks.
What, then, are the policy implications that Herrnstein and Murray draw from the findings that they report?
In The Blank Slate: The Modern Denial of Human Nature, cognitive science, linguist and popular science writer Steven Pinker popularizes the notion that recognizing the existence of innate differences between individuals and groups in traits such as intelligence does not necessarily lead to ‘right-wing’ political implications.
Thus, a leftist might accept the existence of innate differences in ability, but conclude that, far from justifying inequality, this is all the more reason to compensate the, if you like, ‘cognitively disadvantaged’ for their innate deficiencies, differences which are, being innate, hardly something for which they can legitimately be blamed.
Herrnstein and Murray reject this conclusion, but acknowledge it is compatible with their data. Thus, in an afterword to later editions, Murray writes:
“If intelligence plays an important role in determining how well one does in life, and intelligence is conferred on a person through a combination of genetic and environmental factors over which that person has no control, the most obvious political implication is that we need a Rawlsian egalitarian state, compensating the less advantaged for the unfair allocation of intellectual gifts” (p554).
Interestingly, Pinker’s notion of a ‘hereditarian left’, and the related concept of ‘Bell Curve liberals’, is not entirely imaginary. On the contrary, it used to be quite mainstream.
Thus, it was the radical leftist post-war Labour government that imposed the tripartite system on schools in the UK in 1945, which involved allocating pupils to different schools on the basis of their performance in what was then called the 11-plus exam, conducted at with children at age eleven, which tested both ability and acquired knowledge. This was thought by leftists to be a fair system that would enable bright, able youngsters from deprived and disadvantaged working-class backgrounds to achieve their full potential.
On the contrary, in advocating, in his famous (plagiarized) aphorism “From each according to his ability; to each according to his need”, Marx implicitly recognized that individuals differ in “ability”, and, given that, in the unrealistic communist utopia he envisaged, environmental conditions were ostensibly to be equalized, these differences he presumably conceived of as innate in origin.
Thus, one might well reasonably contend that the ‘cognitively disadvantaged’ should be compensated for their innate deficiencies through economic redistribution. Indeed, to some extent, most Western polities already do this, by providing welfare payments and state-funded, or state-subsidized, care to those whose cognitive impairment is such as to qualify as a disability and hence render them incapable of looking after or providing for themselves.
However, we are unlikely to believe that such persons should be given entry to medical school such that they are one day liable to be responsible for performing heart surgery on us or diagnosing our medical conditions.
In short, socialist redistribution is defensible – but affirmative action is definitely not!
Reception and Readability
The reception accorded ‘The Bell Curve’ in 1994 echoed that accorded another book that I have also recently reviewed, but that was published some two decades earlier, namely Edward O. Wilson’s Sociobiology: The New Synthesis.
Both were greeted with similar indignant moralistic outrage by many social scientists, who even employed similar pejorative soundbites (‘genetic determinism’, ‘reductionism’, ‘biology as destiny’), in condemning the two books. Moreover, in both cases, the academic uproar even spilled over into a mainstream media ‘moral panic’, with pieces appearing the popular press attacking the two books.
Yet, in both cases, the controversy focused almost exclusively on just a small part of each book – the single chapter in ‘Sociobiology: The New Synthesis’ focusing on humans and the few chapters in ‘The Bell Curve’ discussing race.
In truth, however, both books were massive tomes of which these sections represented only a small part.
Indeed, due to their size, one suspects most critics never actually read the books in full for themselves, including, it seemed, most of those nevertheless taking it upon themselves to write critiques. This is what led to the massive disconnect between what most people thought the books said, and their actual content.
However, there is a crucial difference.
‘Sociobiology: The New Synthesis’ was a long book of necessity, given the scale of the project Wilson set himself.
As I have written in my review of that latter work, the scale of Wilson’s ambition can hardly be exaggerated. He sought to provide a new foundation for the whole field of animal behaviour, then, almost as an afterthought, sought to extend this ‘New Synthesis’ to human behaviour as well, which meant providing a new foundation, not for a single subfield within biology, but for several whole disciplines (psychology, sociology, economics and cultural anthropology) that were formerly almost unconnected to biology. Then, in a few provocative sentences, he even sought to provide a new foundation for moral philosophy, and perhaps epistemology too.
‘Sociobiology: The New Synthesis’ was, then, inevitably and of necessity, a long book. Indeed, given that his musings regarding the human species were largely (but not wholly) restricted to a single chapter, one could even make a case that it was too short – and it is no accident that Wilson subsequently extended his writings with regard to the human species to a book length manuscript.
Yet, while Sociobiology was of necessity a long book, ‘The Bell Curve: Intelligence and Class Structure in America’ is, for me, unnecessarily overlong.
After all, Herrnstein and Murray’s thesis was actually quite simple – namely that cognitive ability, as captured by IQ testing, is a major correlate of many important social outcomes in modern America.
Yet they reiterate this point, for different social outcomes, again and again, chapter after chapter, repeatedly.
In my view, Herrnstein and Murray’s conclusion would have been more effectively transmitted to the audience they presumably sought to reach had they been more succinct in their writing style and presentation of their data.
Had that been the case then perhaps rather more of the many people who bought the book, and helped make it into an unlikely nonfiction bestseller in 1994, might actually have managed to read it – and perhaps even been persuaded by its thesis.
For casual readers interested in this topic, I would recommend instead Intelligence, Race, And Genetics: Conversations With Arthur R. Jensen (which I have reviewed here, here and here).
“The Bell Curve (1994), runs to more than 800 pages but can be summarised in a few sentences. Black people are more stupid than white people: always have been, always will be. This is why they have less economic and social success. Since the fault lies in their genes, they are doomed to be at the bottom of the heap now and forever” (Wheen 2000).
In making this claim, Wheen clearly demonstrates that he has read few if any of those 800 pages to which he refers.
 Although their discussion of the evidence relating to the causes, genetic or environmental, of the black-white test score gap is extensive, it is not exhaustive. For example, Phillipe Rushton, the author of Race Evolution and Behavior (reviewed here and here) argues that, despite the controversy their book provoked, Herrnstein and Murray actually “didn’t go far enough on race”, omitting, for example, any real discussion, save a passing mention in Appendix 5, of race differences in brain size (Rushton 1997). On the other hand, Herrnstein and Murray also did not mention studies that failed to establish any correlation between IQ and blood groups among African-Americans, studies interpreted as supporting an environmentalist interpretation of race differences in intelligence (Loehlin et al 1973; Scarr et al 1977). For readers interested in a more complete discussion of the evidence regarding the relative contributions of environment and heredity to the differences in IQ scores of different races, see my review of Richard Lynn’s Race Differences in Intelligence: An Evolutionary Analysis, available here.
 For example, some of those accused of serious crimes have been accused of deliberately getting questions wrong on IQ tests in order to qualify as mentally subnormal when before the courts for sentencing in order to be granted mitigation of sentence on this ground, or, more specifically, in order to evade the death penalty.
 This may be because whites or Asians with such low IQs are more likely to have such impaired cognitive abilities because of underlying conditions (e.g chromosomal abnormalities, brain damage) that handicap them over and above the deficit reflected in IQ score alone. On the other hand, blacks with similarly low IQs are still within the normal range for their own race. Therefore, rather than suffering from, say, a chromosomal abnormality or brain damage, they are relatively more likely to simply be at the tail-end of the normal range of IQs within their group, and hence normal in other respects.
 The term Rawlsian is a reference to political theorist John Rawles version of social contract theory, whereby he poses the hypothetical question as to what arrangement of political, social and economic affairs humans would favour if placed in what he called ‘the original position’, where they would be unaware of, not only their own race, sex and position in to the socio-economic hierarchy, but also, most important for our purposes, their own level of innate ability. This Rawles referred to as ‘veil of ignorance’.
 The tripartite system did indeed enable many working-class children to achieve a much higher economic status than their parents, although this was partly due to the expansion of the middle-class sector of the economy over the same time-period. It was also later Labour administrations who largely abolished the 11-plus system, not least because, unsurprisingly given the heritability of intelligence and personality, children from middle-class backgrounds tended to do better on it than did children from working-class backgrounds.
Barrick & Mount 1991 The big five personality dimensions and job performance: a meta-analysis. Personnel Psychology 44(1):1–26
Beckford (2008) Working classes ‘lack intelligence to be doctors’, claims academic, Daily Telegraph, 04 Jun 2008.
Clark 2008 Higher social classes have significantly HIGHER IQs than working class, claims academic Daily Mail, 22 May 2008.
Kruger & Dunning (1999) Unskilled and Unaware of It: How Difficulties in Recognizing One’s Own Incompetence Lead to Inflated Self-Assessments. Journal of Personality and Social Psychology 77(6):1121-34
Loehlin et al (1973) Blood group genes and negro-white ability differences. Behavior Genetics 3(3): 263-270
Rushton, J. P. (1997). Why The Bell Curve didn’t go far enough on race. In E. White (Ed.), Intelligence, political inequality, and public policy (pp. 119-140). Westport, CT: Praeger.
Scarr et al (1977) Absence of a relationship between degree of white ancestry and intellectual skills within a black population. Human Genetics 39(1):69-86 .
Wheen (2000) The ‘science’ behind racism. Guardian, 10 May 2000.